Posters

Abstract

Accurate hourly production time series are essential for modelling the operational behaviour and reliability of hybrid renewable energy plants. Conventional approaches often apply aggregated percentage losses, failing to capture the sequential and compounding effects of different loss mechanisms on an hourly basis. This study presents a deterministic, category‑based framework for modelling wind generation losses with full temporal fidelity, enabling direct integration with hybrid performance simulations. Loss mechanisms are structured into six sequential, non‑overlapping categories: 1. Wind losses - wake and blockage effects reducing the available wind resource 2. Aerodynamic losses - losses occurring below rated power, such as sub-optimal power curve performance and aerodynamic degradation. 3. Control losses – operational scheduling/ curtailment due to noise, shadow flicker and other environmental mitigations, and wind sector management. 4. Electrical losses – transmission and conversion inefficiencies. 5. Export losses – grid capacity and offtaker constraints, determined through co-optimised hybrid modelling with solar PV and energy storage. 6. Stochastic losses – turbine/ system availability and unforeseen outages, incorporated within financial model unless dedicated reliability and risk modelling is required. We propose a simplified wind speed reduction approach to consolidate Category 2 aerodynamic losses, preserving sequential application logic and reducing bias from lumped assumptions. Export constraints (Category 5) are resolved using dispatch level hybrid simulations to reflect real world curtailment dynamics.  Stochastic losses (Category 6) are applied as the final step to avoid underestimating export related curtailments, considering the random nature of unavailability events; In many cases these losses may be represented as an aggregated input to the financial model so that the progression from gross yield to system level net yield represents the P50 scenario without randomness introduced by unavailability events. Where system reliability or power‑market uncertainty is critical, stochastic modelling is recommended (e.g., a Monte-Carlo process) to account for the number of turbines, the distribution of individual turbine unavailability and the targeted overall unavailability level. This methodology offers a transparent, replicable framework for hybrid resource assessment. Application to utility scale hybrid datasets demonstrated improved system-level energy prediction accuracy of 4–6% compared to aggregated loss methods, with negligible computational overhead. The approach can be standardised across projects to support robust grid integration studies, levelised cost of energy (LCOE) calculations, and investment decisions.